Grant-in-Aid for Transformative Research Areas (A) Analysis and synthesis of deep SHITSUKAN information in the real world


D02-8 Roles of material shitsukan on affective shitsukan examined by deep neural networks and EEG


Takehiro Nagai Department of Information and Communications Engineering, School of Engineering, Tokyo Institute of Technology

What is the role of material perception in humans? Shitsukan features can be roughly divided into two categories: material and affective. In affective engineering, it is commonly assumed that affective shitsukan is hierarchically based on material shitsukan for understanding affective recognition. However, it is also plausible that affective shitsukan, such as aesthetics, arises directly from low-order image features instead of material shitsukan perception, such as glossiness. Investigating this possibility requires more than just static psychological experiments like the SD method. This study aims to elucidate the causal relationship between material and affective shitsukan from multiple perspectives. Specifically, we examine the neural hierarchy between various shitsukan features using EEG analysis and investigate the hierarchy of shitsukan representations in deep neural networks. Additionally, we explore the causal relationships between different shitsukan features through multimodal psychophysical experiments in which material shitsukan perception is manipulated.